Increased intratumoral infiltration in IDH wild-type lower-grade gliomas observed with diffusion tensor imaging.

PURPOSE: Diffuse lower grade gliomas (LGG) with isocitrate dehydrogenase (IDH) gene mutations (IDHMUT) have a distinct survival advantage compared with IDH wild-type (IDHWT) cases but the mechanism underlying this disparity is not well understood. Diffusion Tensor Imaging (DTI) has identified infiltrated non-enhancing tumor regions that are characterized by low isotropic (p) and high anisotropic (q) diffusion tensor components that associate with poor survival in glioblastoma. We hypothesized that similar regions are more prevalent in IDHWT (vs. IDHMUT) LGG. METHODS: p and q maps were reconstructed from preoperative DTI scans in N = 41 LGG patients with known IDH mutation and 1p/19q codeletion status. Enhancing and non-enhancing tumor volumes were autosegmented from standard (non-DTI) MRI scans. Percentage non-enhancing tumor volumes exhibiting low p and high q (Vinf) were then determined using threshold values (p = 2 × 10-3mm2/s, q = 3 × 10-4 mm2/s) and compared between IDHWT and IDHMUT LGG, and between IDHMUT LGG with and without 1p/19q codeletion. RESULTS: Vinf volumes were significantly larger in IDHWT LGG than in IDHMUT LGG (35.4 ± 18.3% vs. 15.9 ± 7.6%, P < 0.001). Vinf volumes did not significantly differ between IDHMUT LGG with and without 1p/19q codeletion (17.1 ± 9.5% vs. 14.8 ± 5.8%, P = 1.0). CONCLUSION: IDHWT LGG exhibited larger volumes with suppressed isotropic diffusion (p) and high anisotropic diffusion (q) which reflects regions with increased cell density but non-disrupted neuronal structures. This may indicate a greater prevalence of infiltrative tumor in IDHWT LGG.

Imaging characteristics of cervical spine extra-arachnoid fluid collections managed conservatively.

OBJECTIVE: Determine the MRI characteristics of large post-traumatic cervical spine extra-arachnoid collections managed conservatively in clinically stable patients and whether evidence of clinical or imaging deterioration materialized. MATERIALS AND METHODS: Following IRB approval, we conducted a retrospective search for all patients (>16 years old) over a 17-months period who had an extra-arachnoid fluid collection reported on a post-traumatic cervical spine MRI. Patients were excluded if they had surgery for an unstable fracture (n = 21), emergent decompression (n = 1) or lacked a follow-up MRI within 15 days (n = 1). Two MSK radiologists recorded the size, morphology and MRI signal characteristics of the collections. RESULTS: Eight patients (5 male, 3 female) met the inclusion criteria (mean age 40 years; range 19-78 years). Seven of the eight patients had fluid collections that demonstrated thin, tapered margins, extended >7 vertebral bodies and involved >180 degrees of the spinal canal. The signal characteristics of these collections varied: hyper-T1/iso-T2 (n = 1), iso-T1/T2 (n = 3), hyper-T1/hypo-T2 (n = 3) and mixed-T1/T2 (n = 1). Six of seven collections were ventral. Follow-up MRI demonstrated resolution/significant decrease in size (n = 4 between 1 and 12 days) or no change/slight decrease in size (n = 3; between 2 and 11 days). None of the seven fluid collections enlarged, no patient had abnormal cord signal, and no patient's neurologic symptoms worsened. One of eight patients had a dorsal "mass-like" collection that was slightly smaller 9 days later. CONCLUSION: In stable patients with large, tapered post-traumatic cervical spine extra-arachnoid collections managed non-surgically, none developed (1) clinical worsening, (2) abnormal cord signal or (3) collection enlargement, regardless of the collection's signal characteristics.